Some Disaster Recovery (DR) solutions involve storage replication using an asymmetric ‘active/active’ setup in which components at both the source and target sites are active. An example includes primary (source) and secondary (target) virtual machines that are active and online in the DR environment with at least a portion of storage of the replication source being replicated to the replication target. The use of active/active DR can help achieve cost effective DR in a cloud environment since some compute capacity of a DR-involved virtual machine is used for running other workloads when not in a failover situation. However, enhancements are needed for DR solutions and particularly with respect to orchestration of seamless failover to the secondary site in the event that the primary site experiences a catastrophic failure.